search

roboflow / supervision

We write your reusable computer vision tools. πŸ’œ
https://supervision.roboflow.com
MIT License
24.06k stars 1.79k forks source link

Can use as_yolo convert coco2yolo to segmentation data format ? #615

Open lonngxiang opened 11 months ago

lonngxiang commented 11 months ago

Search before asking

Bug

use this script the conversion format is incorrect https://docs.ultralytics.com/datasets/segment/coco/#dataset-yaml

import supervision as sv

sv.DetectionDataset.from_coco(
    images_directory_path= r"C:\Users\loong\Downloads\Car\valid",
    annotations_path=r"C:\Users\loong\Downloads\Car\valid\_annotations.coco.json",
    force_masks=True
).as_yolo(
    images_directory_path=r"C:\Users\loong\Downloads\Car_yolo\val\images",
    annotations_directory_path=r"C:\Users\loong\Downloads\Car_yolo\val\labels",
    data_yaml_path=r"C:\Users\loong\Downloads\Car_yolo\data.yaml"
)

Environment

No response

Minimal Reproducible Example

No response

Additional

No response

Are you willing to submit a PR?

SkalskiP commented 11 months ago

Hi @lonngxiang πŸ‘‹πŸ» Thanks for your interest in Supervision. Could you be a bit more specific? What do you mean by incorrect?

lonngxiang commented 11 months ago

Hi @lonngxiang πŸ‘‹πŸ» Thanks for your interest in Supervision. Could you be a bit more specific? What do you mean by incorrect?

yeap,I want to use Ultralytics' YOLOv8 to train a segmentation algorithm, but the provided function seems to convert COCO dataset into YOLO format. Could you please help me how to convert it into the required format for segmentation dataset?

Please refer to for details: https://github.com/roboflow/supervision/issues/267#issuecomment-1821974837

lonngxiang commented 11 months ago

Hi @lonngxiang πŸ‘‹πŸ» Thanks for your interest in Supervision. Could you be a bit more specific? What do you mean by incorrect?

Using this script looks just converted to yolo object detection format, not segmentation data set format requirements

import supervision as sv

sv.DetectionDataset.from_coco(
    images_directory_path= r"C:\Users\loong\Downloads\Car\valid",
    annotations_path=r"C:\Users\loong\Downloads\Car\valid\_annotations.coco.json",
    force_masks=True
).as_yolo(
    images_directory_path=r"C:\Users\loong\Downloads\Car_yolo\val\images",
    annotations_directory_path=r"C:\Users\loong\Downloads\Car_yolo\val\labels",
    data_yaml_path=r"C:\Users\loong\Downloads\Car_yolo\data.yaml"
)
SkalskiP commented 11 months ago

@lonngxiang so if you examine the txt file that is produced as a result of conversion it contain box coordinates not segmentation coordinates? Could you paste here content of one of those files?

lonngxiang commented 11 months ago

@lonngxiang so if you examine the txt file that is produced as a result of conversion it contain box coordinates not segmentation coordinates? Could you paste here content of one of those files?

Yes, the conversion of YAML and labels TXT files do not meet the segmentation dataset format requirements

yaml file：


nc: 2
names: [Car, Car]

labes txt file :

1 0.50625 0.24844 0.50469 0.25000 0.50000 0.25000 0.49844 0.25156 0.49375 0.25156 0.49219 0.25312 0.48750 0.25312 0.48594 0.25469 0.48125 0.25469 0.47969 0.25625 0.47500 0.25625 0.47344 0.25781 0.47031 0.25781 0.46875 0.25938 0.46406 0.25938 0.46250 0.26094 0.45781 0.26094 0.45625 0.26250 0.45156 0.26250 0.45000 0.26406 0.44531 0.26406 0.44375 0.26562 0.43906 0.26562 0.43750 0.26719 0.43281 0.26719 0.43125 0.26875 0.42969 0.26875 0.42500 0.27344 0.42344 0.27344 0.41406 0.28281 0.41250 0.28281 0.40469 0.29063 0.40313 0.29063 0.38906 0.30469 0.38906 0.30625 0.37188 0.32344 0.37188 0.32500 0.35469 0.34219 0.35469 0.34375 0.33750 0.36094 0.33750 0.36250 0.32812 0.37188 0.32500 0.36875 0.32344 0.36875 0.32031 0.36562 0.31875 0.36562 0.31719 0.36719 0.31406 0.36719 0.31250 0.36875 0.31094 0.36875 0.30938 0.37031 0.30938 0.37188 0.30781 0.37344 0.30938 0.37500 0.31094 0.37500 0.31250 0.37656 0.31406 0.37656 0.31562 0.37812 0.31562 0.38125 0.31406 0.38281 0.31406 0.39062 0.31250 0.39219 0.31250 0.39375 0.30938 0.39687 0.30781 0.39687 0.30625 0.39844 0.30312 0.39844 0.30156 0.40000 0.30000 0.40000 0.29844 0.40156 0.29531 0.40156 0.29375 0.40313 0.29219 0.40313 0.29063 0.40469 0.28750 0.40469 0.28594 0.40625 0.28437 0.40625 0.28281 0.40781 0.27969 0.40781 0.27813 0.40937 0.27656 0.40937 0.27500 0.41094 0.27187 0.41094 0.27031 0.41250 0.26875 0.41250 0.26719 0.41406 0.26406 0.41406 0.26250 0.41563 0.26094 0.41563 0.25938 0.41719 0.25625 0.41719 0.25469 0.41875 0.25312 0.41875 0.25156 0.42031 0.24844 0.42031 0.24688 0.42188 0.24531 0.42188 0.24375 0.42344 0.24063 0.42344 0.23906 0.42500 0.23750 0.42500 0.23594 0.42656 0.23281 0.42656 0.23125 0.42812 0.22969 0.42812 0.22812 0.42969 0.22500 0.42969 0.22344 0.43125 0.22187 0.43125 0.22031 0.43281 0.21875 0.43281 0.21406 0.43750 0.21250 0.43750 0.20312 0.44688 0.20156 0.44688 0.19375 0.45469 0.19219 0.45469 0.18594 0.46094 0.18594 0.46250 0.18281 0.46563 0.18281 0.46719 0.17969 0.47031 0.17969 0.47187 0.17656 0.47500 0.17656 0.47656 0.17344 0.47969 0.17344 0.48125 0.17031 0.48438 0.17031 0.48594 0.16719 0.48906 0.16719 0.49062 0.16562 0.49219 0.16562 0.49375 0.16406 0.49531 0.16406 0.49844 0.16250 0.50000 0.16250 0.50156 0.16094 0.50313 0.16094 0.50625 0.15937 0.50781 0.15937 0.51094 0.15781 0.51250 0.15781 0.51562 0.15625 0.51719 0.15625 0.52031 0.15469 0.52187 0.15469 0.52344 0.15313 0.52500 0.15313 0.52812 0.15156 0.52969 0.15156 0.53281 0.15000 0.53438 0.15000 0.53750 0.14844 0.53906 0.14844 0.54219 0.14688 0.54375 0.14688 0.54531 0.14531 0.54688 0.14531 0.55000 0.14375 0.55156 0.14375 0.55469 0.14219 0.55625 0.14219 0.59844 0.14375 0.60000 0.14375 0.64375 0.14219 0.64531 0.14219 0.65469 0.14062 0.65625 0.14062 0.66563 0.13906 0.66719 0.13906 0.67656 0.13750 0.67812 0.13750 0.68281 0.13906 0.68437 0.14062 0.68437 0.14531 0.68906 0.14688 0.68906 0.15156 0.69375 0.15313 0.69375 0.15625 0.69688 0.16250 0.69688 0.16406 0.69844 0.17656 0.69844 0.17813 0.70000 0.19062 0.70000 0.19219 0.70156 0.20469 0.70156 0.20625 0.70312 0.21719 0.70312 0.21875 0.70469 0.23125 0.70469 0.23281 0.70625 0.24531 0.70625 0.24688 0.70781 0.25938 0.70781 0.26094 0.70937 0.27344 0.70937 0.27500 0.71094 0.28750 0.71094 0.28906 0.71250 0.30156 0.71250 0.30312 0.71406 0.31562 0.71406 0.31719 0.71562 0.32969 0.71562 0.33125 0.71719 0.34219 0.71719 0.34375 0.71875 0.35625 0.71875 0.35781 0.72031 0.37031 0.72031 0.37188 0.72188 0.38438 0.72188 0.38594 0.72344 0.39531 0.72344 0.39687 0.72500 0.40313 0.72500 0.40469 0.72656 0.41094 0.72656 0.41250 0.72813 0.41875 0.72813 0.42031 0.72969 0.42656 0.72969 0.42812 0.73125 0.43438 0.73125 0.43594 0.73281 0.44219 0.73281 0.44375 0.73438 0.46563 0.73438 0.46719 0.73281 0.50313 0.73281 0.50469 0.73125 0.52500 0.73125 0.52656 0.72969 0.53281 0.72969 0.53438 0.72813 0.54063 0.72813 0.54219 0.72656 0.55000 0.72656 0.55156 0.72500 0.55781 0.72500 0.55937 0.72344 0.56563 0.72344 0.56719 0.72188 0.57344 0.72188 0.57500 0.72031 0.57812 0.72031 0.57969 0.72188 0.57969 0.72344 0.58281 0.72656 0.58281 0.72813 0.58594 0.73125 0.58594 0.73281 0.58906 0.73594 0.58906 0.73750 0.59219 0.74062 0.59219 0.74219 0.59375 0.74375 0.60313 0.74375 0.60469 0.74219 0.61406 0.74219 0.61875 0.73750 0.61875 0.73594 0.62500 0.72969 0.62500 0.72813 0.62656 0.72656 0.62656 0.72344 0.62813 0.72188 0.62813 0.72031 0.62969 0.71875 0.62969 0.71562 0.63125 0.71406 0.63125 0.71094 0.63281 0.70937 0.63281 0.70625 0.63438 0.70469 0.63438 0.70156 0.63594 0.70000 0.63594 0.69844 0.63750 0.69688 0.63750 0.69375 0.63906 0.69219 0.63906 0.69063 0.64062 0.68906 0.64219 0.68906 0.64375 0.68750 0.64531 0.68750 0.64844 0.68437 0.65000 0.68437 0.65312 0.68125 0.65469 0.68125 0.65625 0.67969 0.65781 0.67969 0.66094 0.67656 0.66250 0.67656 0.66563 0.67344 0.66719 0.67344 0.66875 0.67188 0.67031 0.67188 0.67188 0.67031 0.67344 0.67031 0.67500 0.66875 0.67656 0.66875 0.67812 0.66719 0.67969 0.66719 0.68125 0.66563 0.68437 0.66563 0.68594 0.66406 0.68750 0.66406 0.68906 0.66250 0.69219 0.66250 0.69375 0.66094 0.69531 0.66094 0.69688 0.65938 0.69844 0.65938 0.70000 0.65781 0.70312 0.65781 0.70469 0.65625 0.70625 0.65625 0.70781 0.65469 0.70937 0.65469 0.71094 0.65312 0.71406 0.65312 0.71562 0.65156 0.71719 0.65156 0.71875 0.65000 0.72031 0.65000 0.72188 0.64844 0.72500 0.64844 0.72656 0.64687 0.72813 0.64687 0.72969 0.64531 0.73281 0.64531 0.73438 0.64375 0.73594 0.64375 0.73750 0.64219 0.73906 0.64219 0.74062 0.64062 0.74687 0.64062 0.74844 0.63906 0.75625 0.63906 0.75781 0.63750 0.76250 0.63750 0.76406 0.63906 0.76562 0.63906 0.76719 0.64062 0.76875 0.64062 0.77187 0.64375 0.77344 0.64375 0.77656 0.64687 0.77812 0.64687 0.77969 0.64844 0.78125 0.64844 0.78281 0.65000 0.78594 0.65000 0.78750 0.64844 0.79219 0.64844 0.79375 0.64687 0.79531 0.64687 0.79688 0.64531 0.79688 0.64375 0.79844 0.64219 0.79844 0.64062 0.80156 0.63750 0.80156 0.63594 0.80312 0.63438 0.80312 0.63281 0.80469 0.63125 0.80469 0.62813 0.80625 0.62656 0.80625 0.61875 0.80781 0.61719 0.80781 0.61094 0.80937 0.60938 0.80937 0.60313 0.81094 0.60156 0.81094 0.59375 0.81250 0.59219 0.81250 0.58750 0.81406 0.58594 0.81406 0.58437 0.81563 0.58281 0.81563 0.58125 0.81719 0.57969 0.81719 0.57812 0.81875 0.57656 0.81875 0.57500 0.82031 0.57344 0.82031 0.44531 0.81875 0.44375 0.81875 0.43906 0.81719 0.43750 0.81719 0.43281 0.81563 0.43125 0.81563 0.42656 0.81406 0.42500 0.81406 0.42031 0.81250 0.41875 0.81250 0.41406 0.81094 0.41250 0.81094 0.40781 0.80937 0.40625 0.80937 0.40156 0.80781 0.40000 0.80781 0.39531 0.80625 0.39375 0.80625 0.38906 0.80469 0.38750 0.80469 0.38438 0.80312 0.38281 0.80156 0.38281 0.79844 0.37969 0.79688 0.37969 0.79531 0.37812 0.79375 0.37812 0.79063 0.37500 0.78906 0.37500 0.78594 0.37188 0.78438 0.37188 0.78125 0.36875 0.77969 0.36875 0.77812 0.36719 0.77656 0.36719 0.77344 0.36406 0.77187 0.36406 0.76875 0.36094 0.76875 0.35938 0.76562 0.35625 0.76562 0.35469 0.76250 0.35156 0.76250 0.35000 0.75938 0.34687 0.75938 0.34531 0.75625 0.34219 0.75625 0.34063 0.75313 0.33750 0.75313 0.33594 0.74844 0.33125 0.74844 0.32969 0.74531 0.32656 0.74531 0.32500 0.74219 0.32188 0.74219 0.32031 0.73906 0.31719 0.73906 0.31562 0.73594 0.31250 0.73594 0.31094 0.73281 0.30781 0.73281 0.30625 0.69219 0.26562 0.68906 0.26562 0.68750 0.26406 0.67969 0.26406 0.67812 0.26250 0.67031 0.26250 0.66875 0.26094 0.66094 0.26094 0.65938 0.25938 0.65312 0.25938 0.65156 0.25781 0.64375 0.25781 0.64219 0.25625 0.63438 0.25625 0.63281 0.25469 0.62656 0.25469 0.62500 0.25312 0.61719 0.25312 0.61562 0.25156 0.60781 0.25156 0.60625 0.25000 0.59844 0.25000 0.59688 0.24844
1 0.74375 0.31719 0.74531 0.31875 0.74531 0.32031 0.74844 0.32344 0.74844 0.32500 0.75156 0.32812 0.75156 0.32969 0.75469 0.33281 0.75469 0.33437 0.75781 0.33750 0.75781 0.33906 0.76094 0.34219 0.76094 0.34375 0.76250 0.34531 0.76250 0.34687 0.76562 0.35000 0.76562 0.35156 0.76875 0.35469 0.76875 0.35625 0.77187 0.35938 0.77187 0.36094 0.77500 0.36406 0.77500 0.36562 0.77812 0.36875 0.77812 0.37031 0.77969 0.37188 0.78594 0.37188 0.78750 0.37344 0.79375 0.37344 0.79375 0.37188 0.79219 0.37031 0.79219 0.36719 0.79063 0.36562 0.79063 0.36250 0.78906 0.36094 0.78906 0.35781 0.78750 0.35625 0.78750 0.35313 0.78594 0.35156 0.78594 0.35000 0.78438 0.34844 0.78438 0.34531 0.78281 0.34375 0.78281 0.34063 0.78125 0.33906 0.78125 0.33594 0.77969 0.33437 0.77969 0.33125 0.77812 0.32969 0.77812 0.32812 0.77656 0.32812 0.77500 0.32656 0.77187 0.32656 0.77031 0.32500 0.76719 0.32500 0.76562 0.32344 0.76250 0.32344 0.76094 0.32188 0.75625 0.32188 0.75469 0.32031 0.75156 0.32031 0.75000 0.31875 0.74687 0.31875 0.74531 0.31719

image

SkalskiP commented 5 months ago

Hi @lonngxiang πŸ‘‹πŸ» is this still a problem?